Modeling and Under-actuated Control of Stabilization Before Take-off Phase for Flapping-wing Robots

This work studies a stabilization problem of flapping-wing flying robots (FWFRs) before a take-off phase while a robot is on a branch. The claw of the FWFR grasps the branch with enough friction to hold the system steady in a stationary condition. Before the take-off, the claw opens itself and the friction between the claw and branch vanishes. At that moment, the mechanical model turns into an under-actuated multi-link (serial configuration) robotic system where the first joint can rotate freely without any friction as opposed to rotation. The stabilization and balancing are the crucial tasks before take-off. This work explores a new methodology to control an under-actuated lightweight manipulator for its future adaptation to FWFR to improve the stabilization performance before take-off. The setup tries to mimic the birds with two-link legs, a body link, and 2-DoF (degrees of freedom) arms, being all active links except the first passive one. In contrast to common arms, the lightweight-design restriction limits the frame size and requires micromotors. With all of these constraints, control design is a challenge, hence, the system is categorized: a) the leg subsystem (under-actuated), including the two first links, and b) the body and arm subsystem (fully actuated) with the rest of links. The fully-actuated links are controlled by feedback linearization and the under-actuated part with active disturbance rejection control (ADRC) for estimation and rejection of the coupling between both subsystems. The mechanical design, modeling, and control of the proposed system are reported in this work. Experimental results have been also proposed to present a proof of concept for this modeling and control approach.

Automated summary

This work focuses on studying the stability problem of flapping-wing flying robots (FWFRs) before take-off while the robot is on a branch. The claw of the FWFR grasps the branch with enough friction, but opens itself before take-off, leading to a change in the mechanical model. The research explores a new methodology to improve the stabilization performance before take-off by controlling an under-actuated lightweight manipulator.